Synthesis and Characterization of [Ir(1,5-Cyclooctadiene)(μ-H)]4: ATetrametallic Ir4H4-Core, Coordinatively Unsaturated ClusterKuang-Hway Yih,† Isil K. Hamdemir,‡ Joseph E. Mondloch,‡ Ercan Bayram,‡ Saim Ozkar,§ Relja Vasic,⊥

Anatoly I. Frenkel,⊥ Oren P. Anderson,‡ and Richard G. Finke*,‡

†Department of Applied Cosmetology, Hungkuang University, Taichung, Taiwan 433‡Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523, United States§Department of Chemistry, Middle East Technical University, 06531 Ankara, Turkey⊥Department of Physics, Yeshiva University, New York, New York 10016, United States

*S Supporting Information

ABSTRACT: Reported herein is the synthesis of thepreviously unknown [Ir(1,5-COD)(μ-H)]4 (where 1,5-COD= 1,5-cyclooctadiene), from commercially available [Ir(1,5-COD)Cl]2 and LiBEt3H in the presence of excess 1,5-COD in78% initial, and 55% recrystallized, yield plus its unequivocalcharacterization via single-crystal X-ray diffraction (XRD),X-ray absorption fine structure (XAFS) spectroscopy, electro-spray/atmospheric pressure chemical ionization mass spec-trometry (ESI-MS), and UV−vis, IR, and nuclear magneticresonance (NMR) spectroscopies. The resultant productparallelsbut the successful synthesis is different from, videinfrathat of the known and valuable Rh congener precatalyst and synthon, [Rh(1,5-COD)(μ-H)]4. Extensive characterizationreveals that a black crystal of [Ir(1,5-COD)(μ-H)]4 is composed of a distorted tetrahedral, D2d symmetry Ir4 core with two long[2.90728(17) and 2.91138(17) Å] and four short Ir−Ir [2.78680 (12)−2.78798(12) Å] bond distances. One 1,5-COD and twoedge-bridging hydrides are bound to each Ir atom; the Ir−H−Ir span the shorter Ir−Ir bond distances. XAFS provides excellentagreement with the XRD-obtained Ir4-core structure, results which provide both considerable confidence in the XAFSmethodology and set the stage for future XAFS in applications employing this Ir4H4 and related tetranuclear clusters. The [Ir(1,5-COD)(μ-H)]4 complex is of interest for at least five reasons, as detailed in the Conclusions section.

■ INTRODUCTIONMolecular metal clusters1 containing four metal atoms, M4, arean interesting, increasingly important, and evolving area ofinorganic, organometallic, catalytic, and related sciences.Tetrametallic clusters of Ru, Os, Rh, or Ir have beensynthesized, fully characterized, and used as precatalysts forthe catalytic hydrogenation of alkenes, arenes, CO, aldehydes,and ketones as well as in hydroformylation, cyclooligomeriza-tion, cyclization, and polymerization reactions.2 Knowntetrametallic complexes of group 9 metals including Rh, Ir,and Co, and which contain a [M(μ-H)]4 core, include [Rh(1,5-COD)(μ-H)]4 (where 1,5-COD = 1,5-cyclooctadiene),3 [Ir-(CO)(μ-H)H(PPh3)]4,

4 [Ir(η5-C5Me5)(μ-H)]4(BF4)2,5 [Co-

(η5-C5H5)(μ-H)]4,6 and [Co(η5-C5Me4Et)(μ-H)]4.

7

From the above known [M(μ-H)]4-core complexes, thetetrarhodium complex [Rh(1,5-COD)(μ-H)]4 is particularlyrelevant to the present work. This [Rh(1,5-COD)(μ-H)]4complex was first synthesized by Muetterties and co-workersstarting with the commercially available [Rh(1,5-COD)(μ-Cl)]2 complex and K[HB(O-i-Pr)3], as detailed in Scheme 1.3

A single-crystal structural investigation revealed the Rh4 core,with two long and four short Rh−Rh distances. In that

publication, hydrides were located using difference Fouriertechniques between two Rh atoms connected by short Rh−Rhdistances.3 The assignment of four short bonds to the Rh−H−Rh groups is consistent in a general way with the priorliterature.8 [Rh(1,5-COD)(μ-H)]4 proved to be a goodprecatalyst for hydrogenation of toluene in cyclohexane-d12

9

and hydrogenation of carbon dioxide.9 The active catalystspecies is believed to be Rh metal under toluene hydrogenationconditions. Kinetic studies performed on a catalytic carbondioxide hydrogenation system suggested a neutral rhodium(I)hydride species ([Rh(H)(Ph2P(CH2)4PPh2)]x, where x = 1 or2−4) as the active catalyst.9c

Somewhat surprisingly, the Ir analogue of the above Rhcomplex, [Ir(1,5-COD)(μ-H)]4, has not been previouslydescribed, most likely because attempted syntheses, analogousto that of the Rh congener, fail (yields ≤ 1%, vide infra). The56-total-electron [Ir(1,5-COD)(μ-H)]4 complex has a formal17 electron count at each Ir and thus is coordinativelyunsaturated.

Received: December 8, 2011Published: February 22, 2012

Article

pubs.acs.org/IC

© 2012 American Chemical Society 3186 dx.doi.org/10.1021/ic2026494 | Inorg. Chem. 2012, 51, 3186−3193

Our recent work on Ir-based models of a Ziegler-typeindustrial hydrogenation catalyst prepared from [Ir(1,5-COD)-(μ-O2C8H15)]2 plus AlEt3 revealed that Ir4 species are adominant, initial form of the Ir present.10 Hence, [Ir(1,5-COD)(μ-H)]4 with its Ir4H4 core (vide infra) is of interest as anew precursor for testing the formation and stabilizationmechanisms of such Ziegler-type hydrogenation catalysts.10,11

More specifically, the new complex [Ir(1,5-COD)(μ-H)]4 is ofvalue as a fully compositionally and structurally characterizedIr4 analogue of the, on average, Co4-based, subnanometerclusters identified by X-ray absorption fine structure (XAFS)spectroscopy as a dominant species in Co-based Ziegler-typeindustrial hydrogenation catalysts. Such Ziegler-type industrialhydrogenation catalysts12 are used industrially to producestyrenic block copolymers at a level of ∼1.7 × 105 metric tons/year.13 In addition, [Ir(1,5-COD)(μ-H)]4 is of considerableinterest as a possible Ir−H-containing, tetrametallic Ir4H4intermediate in the nucleation and growth of Ir0n nanoclustersstarting with (COD)Ir+ precatalysts14 and with stabilizers suchas [P2W15Nb3O62]

9−, HPO42−, and AlEt3.

10,11,12,15 Whether ornot such polynuclear metal hydride (M−H)n species are keyintermediates in M0

n nanoparticle formationrather than thepresently assumed M0 intermediatesremains controversial.The availability of precatalysts and possible intermediatesderivable from [Ir(1,5-COD)(μ-H)]4 opens up the possibilityof QEXAFS and other direct-method tests with such discrete,fully characterized, Ir4H4-core complexes.14

Herein, we report (i) the 78% initial, and 55% recrystallized,yield synthesis of the previously unknown [Ir(1,5-COD)(μ-H)]4 starting from commercially available precursor [Ir(1,5-COD)(μ-Cl)]2 and LiBEt3H in which added, excess 1,5-COD isone key to the improved yield (vs <1% by the literature routesfor the Rh congener, vide infra), and then (ii) the completecharacterization of the resultant pure, crystalline product bysingle-crystal X-ray diffraction (XRD), XAFS, electrospray/atmospheric pressure chemical ionization mass spectrometry(ESI-MS), UV−vis, IR, and NMR. There are at least fivereasons (a couple of which are given above) as to why thepresent complex is of interest, a full list of which is given as partof the Summary and Possible Future Directions section.

■ EXPERIMENTAL SECTIONMaterials. All manipulations were performed under N2 in a

Vacuum Atmospheres drybox (≤5 ppm O2 as monitored by a VacuumAtmospheres O2-level monitor) or, where noted, using a Schlenk line.All glassware was dried overnight in an oven at 160 °C, cooled undervacuum in a desiccator, and then transferred into the drybox while stillin the desiccator and under vacuum. [Ir(1,5-COD)(μ-Cl)]2 (an orangepowder, Strem Chemicals, 99%), LiBEt3H [as a colorless solution in1.0 M tetrahydrofuran (THF), Aldrich], toluene (Aldrich, 99.8%,anhydrous), and benzene-d6 [Cambridge Isotope Laboratories, Inc.,99.5%, w/o tetramethylsilane (TMS)] were used as received. THF(Mallinckrodt Chemicals AR ACS, 500 mL), n-hexane (Sigma-Aldrich,Reagent Plus, ≥99%, 500 mL), and cyclohexane (Sigma-Aldrich,

99.5%, H2O < 0.001%) were distilled over sodium/benzophenoneunder N2(g) and transferred into the drybox under air-free conditions.Acetone (Burdick and Jackson, >99.9% purity, 0.44% water) and H2O(Nanopure ultrapure H2O system, D4754) were degassed byconnecting to a Schlenk line and then passing Ar for 5 min throughthe solution.

Synthesis of [Ir(1,5-COD)(μ-H)]4. In the drybox, the orangepowdered [Ir(1,5-COD)(μ-Cl)]2 (1.3434 g, 2 mmol) was weighed outand then transferred into a 100 mL round-bottomed Schlenk flaskequipped with a side arm and a 5/8 ×

5/16 in. Teflon-coated magneticstirbar. The flask was sealed, removed from the drybox, and placed ona Schlenk line under Ar via its side arm. Next, 70 mL of roomtemperature THF was added to the flask using a cannula, forming anorange solution with some undissolved orange powder. The flaskcontaining the orange solution of [Ir(1,5-COD)(μ-Cl)]2 was placed inan acetone/dry ice bath at −78 °C and stirred for 15 min. A 2.5 mLgastight syringe was purged three times with Ar using a Schlenk lineand then used to measure out 4 mL (4 mmol) of LiBEt3H. LiBEt3Hwas then added dropwise to the orange [Ir(1,5-COD)(μ-Cl)]2solution under an Ar atmosphere with vigorous stirring. The originalorange color of the solution changed to dark brown upon the dropwiseaddition of LiBEt3H. The resulting solution was stirred at −78 °C foran additional 10 min and then warmed to room temperature. Thesolution slowly turned from dark brown to dark green within 10 minof additional stirring at room temperature. 1,5-COD (12.3 mL, 25equiv per Ir) was measured out with a 20 mL gastight syringe purgedwith Ar and then added over 5−10 min to the dark-green solution.The resulting bright-green solution was stirred at room temperaturefor 24 h and then concentrated to ∼5 mL under vacuum at roomtemperature using a Schlenk line. A visually apparent black powder wasformed in the bright-green solution. The black powder was separatedfrom the bright-green solution under Ar using a Schlenk-waremedium-porosity glass frit of ca. 16 μm pore size. The open end ofthe Schlenk glass frit was sealed by a rubber septum. The black powdercollected on top of the glass frit was washed with degassed H2O (5 mL× 2) and then degassed acetone (5 mL × 3) using a gastight syringethat was previously purged with Ar. The black powder was then driedovernight under vacuum at room temperature, resulting in a blackpowder (0.471 g, 78% yield) that was transported to the drybox andstored in the glass frit sealed from the drybox atmosphere via a rubberseptum.

Crystallization was accomplished by weighing out 92 mg of theblack, powdered [Ir(1,5-COD)(μ-H)]4 in the drybox and transferringit into a 15 mL Schlenk tube. The Schlenk tube was then sealed,removed from the drybox, and placed on a Schlenk line under Ar.Next, 2.0 mL of a room temperature n-pentane/THF (20:1) mixturewas measured out using a gastight syringe and then added to the tube.The stopper in the Schlenk tube containing the black powder andTHF was black-electrical-taped to secure it. The contents of theSchlenk tube were heated to approximately 66 °C using a heat gun,while the taped stopper was held manually to secure it during minimalboiling. The resulting solution was clear, bright-green, andhomogeneous with no observable solid or particulate mass. TheSchlenk tube was then placed in a −20 °C freezer. Black crystals of[Ir(1,5-COD)(μ-H)]4 were obtained in the tube after 4 h at −20 °C.At the end of 4 h, the Schlenk tube containing the crystals wasconnected to a Schlenk line, and the liquid portion was removed usinga cannula. The tube containing black [Ir(1,5-COD)(μ-H)]4 crystals

Scheme 1. Balanced Reaction Stoichiometry and the Reaction Conditions for the Synthesis of [Rh(1,5-COD)(μ-H)]4 (Adaptedfrom Reference 3a)

Inorganic Chemistry Article

dx.doi.org/10.1021/ic2026494 | Inorg. Chem. 2012, 51, 3186−31933187

(64 mg, 55% overall yield) was kept under vacuum overnight. Thecrystalline material was then transported back into the drybox andstored in a 2 mL vial. The [Ir(1,5-COD)(μ-H)]4 complex is air-stablein crystalline form. Anal. Calcd for C32H52Ir4 (mol wt 1205.64 g/mol):C, 31.88; H, 4.35. Found: C, 31.74; H, 4.28. ESI-MS peaks (m/z inDa, assigned ion): 1205.2478 ([C32H51Ir4]

+), 1507.3229([C40H66Ir5]

+). UV−vis peaks (nm) in THF: 476, 626. IR bands(cm−1) as a KBr pellet: 697.90, 766.66, 815.67, 859.46, 866.74, 994.42,1072.04, 1146.95, 1167.72, 1203.64, 1233.36, 1295.73, 1320.87,1423.13, 1437.37, 1469.38, 2818.14, 2867.42, 2907.91, 2936.42,2985.69. 1H NMR in benzene-d6 [δ in ppm (multiplicity, number ofH)]: −2.89 (s, 1), 1.37 (m, 4), 2.09 (m, 4), 4.14 (m, 4). 13C NMR inbenzene-d6 (δ in ppm): 68.64, 33.29.Instrumentation and Sample Preparation. XRD. Single

crystals of [(1,5-COD)Ir(μ-H)]4 suitable for XRD analysis weregrown by recrystallization from 20:1 n-pentane/THF usingthe crystallization procedure detailed above. Diffraction data werecollected at 120 K on a Bruker Kappa Apex II diffractometerequipped with graphite-monochromatic Mo Kα (λ = 0.710 73Å) radiation. A suitable single crystal of [Ir(1,5-COD)(μ-H)]4was mounted on a Cryoloop in Paratone-N oil. Initial latticeparameters were determined from 452 reflections harvestedfrom 36 frames. Cell constants and other pertinent crystallo-graphic information are reported in Tables S3−S7 in theSupporting Information. The raw intensity data were integratedand corrected for Lorentz and polarization effects; anabsorption correction was applied to the data using theprogram SADABS from the Apex II16 software package. Thestructure was solved by direct methods and refined using theSHELXTL17 software package. The non-H atoms were refinedwith anisotropic atomic displacement parameters. H atomsbound to C atoms were included in their idealized positionsand were refined with a riding model using isotropic thermalparameters 1.2 times larger than the Ueq value of the atom towhich they were bonded. The two unique hydride atoms of themolecular core were located straightforwardly in the differenceelectron-density map and were refined with isotropic atomicdisplacement parameters.XAFS Spectroscopy. XAFS experiments were performed at

Beamline X-19A at the National Synchrotron Light Source (NSLS)at Brookhaven National Laboratory. Energy was swept from 150 eVbelow to 1528 eV above the Ir L3-edge (edge energy = 11 215 eV) forthe [Ir(1,5-COD)(μ-H)]4 sample. The X-ray absorption coefficientwas measured in transmission mode by positioning the samplebetween the incident and transmission beam detectors. Ir0 black wasused as a reference for the X-ray energy calibration and data alignment.The Ir0 sample was positioned between the transmission and referencebeam detectors and measured simultaneously with the main sample.The X-ray detectors were gas-filled ionization chambers. A samplesolution of initially crystalline [Ir(1,5-COD)(μ-H)]4 was freshlysynthesized at Colorado State University. The black crystal wastransferred into a 5 mL glass vial in a N2-filled Vacuum Atmospheresdrybox (≤5 ppm O2). The glass vial was then double-sealed under N2gas and transported to the NSLS. At the NSLS, the vial was opened ina N2-filled MBraun glovebox and the sample was prepared by brushinga fine powder uniformly onto an adhesive tape, which was then foldedseveral times to achieve a suitable total thickness for the measurement.Within less than 3 min from removing taple sample from the glovebox,the tape sample was transferred into an airtight cell purged with N2 forXAFS measurements.

The data processing and analysis was performed using the IFEFFITpackage.18 The EXAFS analysis was done by fitting the theoreticalFEFF6 signals to the experimental data in r space. Theoreticalcontributions included only the first (Ir−C) and second (Ir−Ir)nearest neighbors (1NN). The passive electron factor, S0

2, was foundto be 0.84 by fits to the standard Ir0 black and then fixed for furtheranalysis of [Ir(1,5-COD)(μ-H)]4. The parameters describing theelectronic properties (correction to the photoelectron energy origin)and local structure environment (coordination numbers N, bondlengths R, and their mean-squared disorder parameters σ2) around theabsorbing atoms were varied during the fitting. There were a total of13 relevant independent data points and 7 variables in the fit.

■ RESULTS AND DISCUSSION

Initial Controls and Attempted Syntheses Based onthe Literature. Initially, to calibrate our hands, the knowntetrametallic hydride [Rh(1,5-COD)(μ-H)]4 was synthesizedby two different researchers using Muetterties’ originalprocedure,3a or Bonnemann’s slightly revised version3c ofMuetterties’ original procedure (see the Supporting Informa-tion for details).19 Pleasingly, dark-red crystals of the [Rh(1,5-COD)(μ-H)]4 complex were obtained in a 50% yield usingboth procedures in our hands.20

Next, the obvious experiments were performed in which weattempted to prepare [Ir(1,5-COD)(μ-H)]4 using each of threeslightly different procedures published for the [Rh(1,5-COD)(μ-H)]4 analogue by Bonnemann,3c Hampden-Smith,3b

and Muetterties3a (see the Supporting Information for thedetails of these failed syntheses).21 A tiny amount of dark-greenpowder was obtained in all three trials. The trace, dark-green,Ir-product powder, from adapting Muetterties’ Rh-congenerprocedure to the Ir case, was characterized using ESI-MS,NMR, IR, UV−vis, and X-ray photoelectron (XPS) spectros-copies (as detailed in the Supporting Information), results thatencouraged us to pursue the superior synthesis reported herein.However, the yield in each case using the adapted literatureprocedures was extremely low (∼1%)even though we wereable to prepare the [Rh(1,5-COD)(μ-H)]4 congener in 50%yield (that matched the literature 50−60% yields, vide supra)prior to the attempted Ir-congener syntheses and as controlexperiments. Moreover, crystallization attempts failed using thesmall amounts of dark-green, Ir-product powder obtained fromeach of the three Muetterties, Bonnemann, and Hampden-Smith adapted syntheses. Specifically, solutions cooled slowlyfrom room temperature to −76 °C in hexane, acetone, orethanol or cooled from room temperature to 10 °C incyclohexane/dichloromethane (1:1) failed to produce singlecrystals. In addition, dissolving the complex (0.4 mg) inpentane (0.5 mL), adding acetone (pentane:acetone = 1:1 byvolume), and keeping the resultant slightly cloudy solution at−78 °C for 10 h failed to provide single crystals. (The reverseorder of the solvent addition was also tried.) These initialcrystallization studies were undoubtedly limited by the smallamounts of Ir product available from the initial, ∼1% yield

Scheme 2. Balanced Reaction Stoichiometry and Reaction Conditions of the Successful Synthesis of [Ir(1,5-COD)(μ-H)]4

Inorganic Chemistry Article

dx.doi.org/10.1021/ic2026494 | Inorg. Chem. 2012, 51, 3186−31933188

syntheses. Hence, the development of a higher yield synthesisbecame the next order of business.Successful Synthesis and Stoichiometry of Formation

of [(1,5-COD)Ir(μ-H)]4. After some trial and error, thesuccessful synthesis of [Ir(1,5-COD)(μ-H)]4 was discovered,a key of which is the use of excess 1,5-COD that was added basedon the hypothesis that it might stabilize the product. Thesuccessful synthesis is carried out starting with THF solutionsof LiBEt3H

22 and [Ir(1,5-COD)(μ-Cl)]2 at −78 °C (Scheme2). Excess 1,5-COD23 (25 equivs/1 equiv of Ir) is added slowlyover 5−10 min after the main reaction and at roomtemperature, resulting in a solution color change from darkgreen to bright green (the latter being the characteristic color ofsolutions when the black-appearing crystals of [Ir(1,5-COD)(μ-H)]4 are dissolved in THF, for example, vide infra). Theresulting black powder is obtained in 78% yield. Followingcrystallization with a n-pentane/THF (20:1) solution at−20 °C, 55% yield of black, crystalline [Ir(1,5-COD)(μ-H)]4was obtained. The black, crystalline [Ir(1,5-COD)(μ-H)]4complex dissolves in THF and benzene and is also slightlysoluble in diethyl ether, n-pentane, n-hexane, acetone, methanol,and acetonitrile.Single-Crystal X-ray Crystallography Structure. The

single-crystal XRD structure of [Ir(1,5-COD)(μ-H)]4 and theresulting atomic numbering scheme are shown in Figure 1. The

space group is Pbcn, and the lattice constants are a =12.5628(3) Å, b = 18.4647(5) Å, and c = 12.3963(3) Å. The[Ir(1,5-COD)(μ-H)]4 molecule is a diamagnetic, 56-total-electron cluster, with formally 17 electrons at each Ir atom (i.e.,unless one would choose to count the two longer Ir−Ir bonds asIrIr double bonds as one way to achieve 18 electron countsat each Ir). The [Ir(1,5-COD)(μ-H)]4 molecule is composed ofa distorted tetrahedral Ir4 core of D2d geometry. Each Ir centeris bonded to two olefinic groups of one 1,5-COD moiety plustwo edge-bridging (vide infra) hydrides. The resulting Ir4H4

core exhibits S4 geometry. The molecule possesses a crystallo-graphic 2-fold symmetry (i.e., the molecule resides on acrystallographic 2-fold axis that connects the halves of themolecule, Ir1 and Ir1A, for example). Selected bond lengthsand angles are given in Tables 1 and 2, respectively. Two Ir−Irdistances are long [2.90728(17) and 2.91138(17) Å] and fourIr−Ir distances are short [2.78680(12)−2.78798(12) Å].24 A

residual electron-density analysis strongly suggests that thehydrides are located between two Ir atoms (i.e., are edge-bridging hydrides) connected by short Ir−Ir distances. Thehydride positions, from refinement of the hydride atoms usingthe procedure detailed in the Experimental Section, appearreasonable but may be influenced by Fourier termination errorsemanating in the Ir atoms. Hence, a neutron-dif f ractionexperiment is needed to reveal the true positions of the hydridesand is planned. That said, the observed short Ir−Ir distances arewithin the range of those of Ir−Ir bonds containing edge-bridging hydrides.4,25 The longer Ir−Ir distances correspond toIr−Ir bonds without bridging hydrides (Ir1−Ir1A and Ir2−Ir2Ain Figure 1). These long Ir−Ir distances are slightly longer thanthe literature values for singly bonded Ir−Ir distances (2.65−2.73 Å).4,2j The observed Ir−H, Ir−C, and C−C bonddistances in [Ir(1,5-COD)(μ-H)]4 [1.67(3) and 1.82(3) Å,2.116(2)−2.182(2) Å, and 1.407(3)−1.424(3) Å, respectively]are consistent with earlier literature.4,11,26 The Ir−Ir−Ir anglesvary between 58.50 and 62.86°, confirming the distortedtetrahedral shape of the Ir4 core.

27 The H−Ir−H [108.2(16)−108.5(15)°] and C−Ir−C [88.85(9)−96.88(9)°] angles areconsistent with those previously reported for similar com-plexes.4

XAFS Characterization. EXAFS and XANES werecollected for two reasons: first, to test whether a minor Ir5species, detected by ESI-MS early in the characterization of thecrystalline complex (Figure S5 of the Supporting Information),was present in the bulk sample of the crystalline material (i.e., inaddition to the most abundant peak expected for the Ir4 species,vide infra). Or, as we suspected, is the ESI-MS-observed Ir5species actually formed during the ESI-MS process and thus anartifact of the ESI-MS? Second, EXAFS and XANES werecollected on the parent [Ir(1,5-COD)(μ-H)]4 complex becausethese spectroscopiesand, hence, the present studyareexpected to be valuable in providing a baseline/background

Figure 1. Single-crystal X-ray structure and atomic numbering schemefor [Ir(1,5-COD)(μ-H)]4 at 50% probability.

Table 1. Selected Bond Lengths (Å) in a [Ir(1,5-COD)(μ-H)]4 Crystal Obtained by XRD Structural Refinement

bond bond length bond bond length

Ir1−Ir2 2.78680(12) Ir1−C7 2.116(2)Ir1−Ir2A 2.78798(12) Ir1−C8 2.156(2)Ir2−Ir1A 2.78797(12) Ir2−C11 2.155(2)Ir1−Ir1A 2.90728(17) Ir2−C12 2.186(2)Ir2−Ir2A 2.91138(17) Ir2−C15 2.128(2)Ir1A−Ir2A 2.78680(12) Ir2−C16 2.158(2)Ir1−H1 1.71(3) C3−C4 1.407(3)Ir1−H2 1.82(3) C7−C8 1.424(3)Ir2−H2 1.67(3) C11−C12 1.405(3)Ir1−C3 2.156(2) C15−C16 1.419(3)Ir1−C4 2.187(2)

Table 2. Selected Bond Angles (deg) in a [Ir(1,5-COD)(μ-H)]4 Crystal

bond bond angle bond bond angle

Ir1−Ir2−Ir2A 58.537(3) H1−Ir1−H2 108.5(15)Ir2−Ir1−Ir1A 58.586(3) H2−Ir2−H1A 108.2(16)Ir2−Ir1−Ir2A 62.966(4) C3−Ir1−C7 96.88(9)Ir2A−Ir1−Ir1A 58.547(3) C4−Ir1−C8 88.85(9)Ir1−Ir2−Ir1A 62.867(4) C11−Ir2−C15 96.77(9)Ir1A−Ir2−Ir2A 58.497(3) C12−Ir2−C16 87.98(9)

Inorganic Chemistry Article

dx.doi.org/10.1021/ic2026494 | Inorg. Chem. 2012, 51, 3186−31933189

study for XAFS characterization of this previously unknowncomplex in future applications in catalysis and other areas.Hence, a bulk sample of crystalline [Ir(1,5-COD)(μ-H)]4 was

examined by EXAFS and XANES spectroscopies. The EXAFSspectrum was analyzed only in the first nearest-neighbor range.Fourier transform (FT) magnitudes of k3-weighted Ir L3-edgeEXAFS data of the [(1,5-COD)Ir(μ-H)]4 complex, and its fitusing Ir−Ir and Ir−C first nearest-neighbor contributions, areshown in Figure 2. Two distinct peaks (uncorrected for the

photoelectron phase shift) at around 2.5 and 1.7 Å are due tothe Ir−Ir and Ir−C scattering contributions, respectively. Theirreal-space distances are 2.80 ± 0.01 and 2.15 ± 0.01 Å for Ir−Irand Ir−C, respectively. The Ir−Ir coordination number (NIr−Ir)of the [(1,5-COD)Ir(μ-H)]4 complex is 3.0 ± 1.2, as expectedfor an Ir4 core. The Ir−C coordination number (NIr−C) in the[(1,5-COD)Ir(μ-H)]4 crystal is 4.2 ± 0.7, as expected from oneCOD attachment to each Ir center. The Ir−Ir and Ir−C bonddistances obtained using EXAFS (2.80 ± 0.01 and 2.15 ± 0.01Å, respectively) are consistent with those determined usingXRD [2.78680(12)−2.91138(17) and 2.116(2)−2.182(2) Å,respectively]. The lack of a higher order contribution beyondthe Ir−Ir scatterer at 2.80 Å attests to the homogeneity of thesamples and the lack of larger Ir clusters.The XANES spectrum of [Ir(1,5-COD)(μ-H)]4 was

obtained and compared to the XANES of both Ir0 black andcrystallographically and EXAFS-characterized10,12 [Ir(1,5-COD)(μ-O2C8H15)]2 (Figure 3). The position and height ofthe main absorption peak (white line) at the Ir L3-edge aresimilar for [Ir(1,5-COD)(μ-H)]4 and Ir0 black samples. On theother hand, the Ir L3-edge white line is shifted to higher energyand reaches higher normalized absorption coefficient values in[Ir(1,5-COD)(μ-O2C8H15)]2 compared to [Ir(1,5-COD)(μ-H)]4, both of which contain formally IrI. This observationindicates a higher positive charge on Ir atoms in the [Ir(1,5-COD)(μ-O2C8H15)]2 complex compared to [Ir(1,5-COD)(μ-H)]4. Restated, the XANES-determined, “effective” oxidationstate of [(1,5-COD)Ir(μ-H)]4 is arguably closer to that of bulkIr0 than the formally IrI in [Ir(1,5-COD)(μ-O2C8H15)]2.However, the overall shape of the XANES spectrum (past thewhite line) of [Ir(1,5-COD)(μ-H)]4 is quite similar to that of[Ir(1,5-COD)(μ-O2C8H15)]2. Both [Ir(1,5-COD)(μ-H)]4 and[Ir(1,5-COD)(μ-O2C8H15)]2 spectra lack the post-edge oscil-latory behavior seen in the bulk Ir0 consistent with the smallcoordination numbers of Ir atoms in both complexes.Additional Characterization Using ESI-MS, UV−vis, IR,

and NMR. ESI-MS28 of the black [Ir(1,5-COD)(μ-H)]4

crystals dissolved in dichloromethane exhibits a most abundantpeak located at 1205.2478 Da (Figures S3−S5 in theSupporting Information). The experimentally observed isotopicpeak distribution pattern matches the simulated isotopicdistribution for [C32H51Ir4]

+, formulated as [Ir4(1,5-COD)4(μ-H)3]

+. The UV−vis spectrum of the dark-greenpowder of [Ir(1,5-COD)(μ-H)]4 dissolved in THF showsabsorption bands at 476 and 626 nm (Figure S6 in theSupporting Information). Of interest is that the experimentalobservation of two absorption maxima at 476 and 626 nm for[Ir(1,5-COD)(μ-H)]4 differ significantly from the UV−visspectra of formally Ir0-containing tetrairidium complexes suchas Ir4(CO)11[tert-butyl-calix[4]arene(OPr)3(OCH2PPh2)] (278and 326 nm),2d or various Ir4(CO)12 clusters (278, 326, and430 nm).29 Furthermore, the UV−vis spectrum of [Ir(1,5-COD)(μ-H)]4 is different from that of formally IrI-containing[Ir(1,5-COD)(μ-O2C8H15)]2 (486 nm), [(1,5-COD)Ir(μ-pyr-azole)]2 (498 and 585 nm), and [(1,5-COD)Ir(μ-6-methyl-2-hydroxypyridine)]2 (484 nm).11,30 In short, computationalassistance will be required before the observed bands at 476and 626 nm in the UV−vis spectrum of [Ir(1,5-COD)(μ-H)]4in THF can be assigned with confidence.The IR spectrum of the [Ir(1,5-COD)(μ-H)]4 complex as

a KBr pellet is, as expected, similar to that of the well-characterized Rh analogue (Figure S7 in the SupportingInformation).3 The 1H NMR spectrum of crystalline [Ir(1,5-COD)(μ-H)]4 dissolved in benzene-d6 (Figure S8 in theSupporting Information) shows a signal at −2.89 ppm and hasthe proper integration for the four Ir−H hydrides.31 The signalsat 4.14, 2.09, and 1.37 ppm are assignable to the olefinic andmethylene H atoms, respectively. The 13C NMR spectrum ofcrystalline [(1,5-COD)Ir(μ-H)]4 dissolved in benzene-d6 showssignals at 68.64 and 33.29 ppm for the COD ligands (Figure S9in the Supporting Information), consistent with the literaturevalues for similar Ir(1,5-COD) complexes (i.e., within theranges of 52.6−92.6 and 27.0−33.4 ppm, respectively).32

Summary and Possible Future Directions. The syn-thesis of the previously unavailable [Ir(1,5-COD)(μ-H)]4complex in 78% initial, and 55% recrystallized, yield wasaccomplished starting with commercially available LiBEt3H and[Ir(1,5-COD)(μ-Cl)]2 in the presence of excess 1,5-COD inTHF. The resultant [Ir(1,5-COD)(μ-H)]4 was fully charac-terized by single-crystal XRD, XAFS, ESI-MS, UV−vis, IR, andNMR. The [Ir(1,5-COD)(μ-H)]4 crystal structure shows a

Figure 2. FT magnitudes of Ir L3-edge EXAFS data for the [(1,5-COD)Ir(μ-H)]4 complex (black) and its associated fit using Ir−C andIr−Ir contributions (red).

Figure 3. XANES of the [Ir(1,5-COD)(μ-H)]4 complex and referencecompounds used of formally Ir0 black and formally IrI in [(1,5-COD)Ir(μ-O2C8H15)]2.

11

Inorganic Chemistry Article

dx.doi.org/10.1021/ic2026494 | Inorg. Chem. 2012, 51, 3186−31933190

distorted tetrahedral, D2d Ir4 core with one 1,5-COD and whatappear to be two edge-bridging hydrides bound to each Ircenter. The Ir−Ir, Ir−H, and Ir−C distances and Ir−Ir−Ir, H−Ir−H, and C−Ir−C bond angles are within the range of thosefor similar complexes from the extant literature. The EXAFS-determined Ir−Ir and Ir−C bond distances are in goodagreement with the XRD results and validate and benchmarkEXAFS as a useful method for characterization of the [Ir(1,5-COD)(μ-H)]4 complex in future applications. The EXAFSresults also are of value in that they demonstrate a high degreeof homogeneity of the bulk [Ir(1,5-COD)(μ-H)]4 sample.As alluded to in the Introduction, there are at least five

reasons why the previously unknown, tetranuclear, coordina-tively unsaturated [Ir(1,5-COD)(μ-H)]4 cluster is of interest,the first of which is (i) that [Ir(1,5-COD)(μ-H)]4 holds thepromise of serving as a multipurpose, coordinatively unsatu-rated, Ir4-based precatalyst and organometallic synthon, analo-gous to its Rh congener [Rh(1,5-COD)(μ-H)]4. Our ownefforts are focused on employing [Ir(1,5-COD)(μ-H)]4 (ii) as aXAFS model/standard and possible Ir4H4 intermediate innucleation and growth studies of Ir0n nanoclusters starting from(1,5-COD)Ir+-based precatalyststhe role of polynuclearMaHb species (M metal), as opposed to just polynuclear M0

nspecies, in nanocluster nucleation and growth being animportant but controversial point at present.14,15 This new,tetrametallic cluster is also of interest (iii) as a discrete, precisecomposition tetrametallic Ir4H4 complex for possible use in thepreparation of both homogeneous and supported, heteroge-neous subnanometer Ir4H4-based catalysts, (iv) as a newprecursor for testing the formation and stabilization mecha-nisms of Ir-based, so-called Ziegler-type industrial hydro-genation model catalysts prepared from (1,5-COD)Ir+-basedprecatalysts and AlEt3,

10 and (v) as a fully compositionally andstructurally characterized Ir4 analogue of the, on average, Co4-based, subnanometer clusters identified by XAFS as a dominantspecies in Co-based, Ziegler-type industrial polymer hydro-genation catalysts.15 Also noteworthy in conclusion is that theCo member of this class of tetranuclear clusters, [M(1,5-COD)(μ-H)]4 (M = Ir, Rh, Co), may be preparable as well,although it remains to be synthesized, isolated, andunequivocally characterized. Hence, it is hoped that the presentsynthesis and characterization, of the previously unavailable[Ir(1,5-COD)(μ-H)]4, will be of value for the above, as well asother, future studies.

■ ASSOCIATED CONTENT

*S Supporting InformationInstrumentation for, and the experimental procedures behind,the ESI-MS, UV−vis, IR, and NMR spectroscopic studies,literature tables for Ir−Ir bond distances and Ir−Ir−Ir bondangles of similar compounds, crystal data and structurerefinement tables with bond distances, bond angles, andanisotropic and isotropic displacement parameters, a CIF filecontaining the crystal structure data, ESI-MS, UV−vis, IR, andNMR spectra, and detailed experimental procedures for (i)successful syntheses of [Rh(1,5-COD)(μ-H)]4, (ii) initial, low-yield-synthesis attempts for [Ir(1,5-COD)(μ-H)]4 whilefollowing literature procedures for the Rh congener, and (iii)control experiments performed to decrease the amount of a 1HNMR-detected impurity. This material is available free ofcharge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*E-mail: rfinke@lamar.colostate.edu.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe thank the following people working as part of the ColoradoState University Central Instrumentation Facility: StephanieFiedler for the XRD data and Donald L. Dick for the ESI-MSdata. Financial support at Colorado State University wasprovided by NSF Grant CHE-0611588 and at YeshivaUniversity by DOE BES Grant DE-FG02-03ER15476. Use ofthe NSLS was supported by the U.S. Department of Energy,Office of Science, Office of Basic Energy Sciences, underContract DE-AC02-98CH10886. Beamline X19A at the NSLSis supported, in part, by the Synchrotron Catalysis Consortium,U.S. Department of Energy, Grant DE-FG02-05ER15688.

■ REFERENCES(1) Muetterties’ 1975 definition of molecular metal clusters is“discrete molecules or molecular ions that contain three or more metalatoms in a bonding interaction”.1a Gates has commented that metalclusters are “compounds with metal−metal bonds stabilized by ligandssuch as carbonyls”.1b A few lead references to the extensive literatureof molecular metal clusters are given below.1a−f Studies comparingstructure, thermochemistry, and bonding interactions in molecularmetal clusters to those of bulk metal surfaces containing chemisorbedligands have revealed that one valuable contribution of molecularmetal clusters is that they can be reasonable approximations to ligated,bulk metal surfaces.1e (a) Muetterties, E. L. Bull. Soc. Chim. Belg. 1975,84, 959−986. (b) Gates, B. C. Angew. Chem., Int. Ed. 1993, 32, 228−229. (c) Chini, P. Pure Appl. Chem. 1970, 23, 489−503. (d) Chini, P.;Longoni, G.; Albano, V. G. Adv. Organomet. Chem. 1976, 14, 285−344.(e) Muetterties, E. L.; Rhodin, T. N.; Band, E.; Brucker, C. F.; Pretzer,W. R. Chem. Rev. 1979, 79, 91−137. (f) Muetterties, E. L. Science1977, 196, 839−848. (g) Chini, P. J. Organomet. Chem. 1980, 200, 37−61. (h) Calabres, J. C.; Dahl, L. F.; Cavalier, A.; Chini, P.; Longoni, G.;Martinen, S. J. Am. Chem. Soc. 1974, 96, 2616−2618. (i) Vranka, R. G.;Dahl, L. F.; Chini, P.; Chatt, J. J. Am. Chem. Soc. 1969, 6, 1574−1576.(j) Morse, M. D. Chem. Rev. 1986, 86, 1049−1109. (k) Adams, R. D.Acc. Chem. Res. 1983, 16, 67−72. (l) Adams, R. D. Chem. Rev. 1989,89, 1703−1712. (m) Adams, R. D. Polyhedron 1985, 4, 2003−2025.(n) Adams, R. D.; Babin, J. E.; Tasi, M. Inorg. Chem. 1986, 25, 4514−4519. (o) Adams, R. D.; Chen, M. Organometallics 2011, 30, 5867−5872. (p) Adams, R. D.; Chen, M. Organometallics 2012, 31, 445−450.(2) (a) Grunwaldt, J.-D.; Kappen, P.; Basini, L.; Clausen, B. S. Catal.Lett. 2002, 78, 13−21. (b) Li, F.; Gates, B. C. J. Phys. Chem. B 2004,108, 11259−11264. (c) Stuntz, G. F.; Shapley, J. R.; Pierpont, C. G.Inorg. Chem. 1978, 17, 2596−2603. (d) Silva, N.; Solovyov, A.; Katz,A. Dalton Trans. 2010, 39, 2194−2197. (e) Goellner, J. F.; Guzman, J.;Gates, B. C. J. Phys. Chem. B 2002, 106, 1229−1238. (f) Li, F.; Gates,B. C. J. Phys. Chem. B 2003, 107, 11589−11596. (g) Uzun, A.; Gates,B. C. Angew. Chem., Int. Ed. 2008, 47, 9245−9248. (h) Xu, Z.; Xiao,F.-S.; Purnell, S. K.; Alexeev, O.; Kawi, S.; Deutsch, S. E.; Gates, B. C.Nature 1994, 372, 346−348. (i) Argo, A. M.; Odzak, J. F.; Lai, F. S.;Gates, B. C. Nature 2002, 415, 623−626. (j) Argo, A. M.; Odzak, J. F.;Gates, B. C. J. Am. Chem. Soc. 2003, 125, 7107−7115. (k) Uzun, A.;Gates, B. C. J. Am. Chem. Soc. 2009, 131, 15887−15894. (l) Doi, Y.;Koshizuka, K.; Keii, T. Inorg. Chem. 1982, 21, 2732−2736. (m) Doi,Y.; Koshizuka, K.; Tamura, S. J. Mol. Catal. 1983, 19, 213−222.(n) Sanchez-Delgado, R. A.; Andriollo, A.; Puga, J.; Martin, G. Inorg.Chem. 1987, 26, 1867−1870. (o) Bradley, J. S. J. Am. Chem. Soc. 1979,101, 7419−7421. (p) Rosas, N.; Marquez, C.; Hernandez, H.; Gomez,R. J. Mol. Catal. 1988, 48, 59−67. (q) Adams, R. D.; Falloon, S. B.Organometallics 1995, 14, 4594−4600.

Inorganic Chemistry Article

dx.doi.org/10.1021/ic2026494 | Inorg. Chem. 2012, 51, 3186−31933191

(3) (a) Kulzick, M.; Price, R. T.; Muetterties, E. L.; Day, V. W.Organometallics 1982, 1, 1256−1258. (b) Duan, Z.; Hampden-Smith,M. J.; Sylwester, A. P. Chem. Mater. 1992, 4, 1146−1148.(c) Bonnemann, H.; Brijoux, W.; Brinkmann, R.; Dinjus, E.;Fretzen, R.; Joußen, T.; Korall., B. J. Mol. Catal. 1992, 74, 323−333.(4) Garlaschelli, L.; Greco, F.; Peli, G.; Manassero, M.; Sansoni, M.;Gobetto, R.; Salassa, L.; Pergola, R. D. Eur. J. Inorg. Chem. 2003,2108−2112.(5) Cabeza, J. A.; Nutton, A.; Mann, B. E.; Brevard, C.; Maitlis, P. M.Inorg. Chim. Acta 1986, 115, L47−L48.(6) Huttner, G.; Lorenz, H. Chem. Ber. 1975, 108, 973.(7) Bau, R.; Ho, N. N.; Schneider, J. J.; Mason, S. A.; McIntyre, G. J.Inorg. Chem. 2004, 43, 555−558.(8) (a) Day, V. W.; Fredrich, M. F.; Reddy, G. S.; Sivak, A. J.; Pretzer,W. R.; Muetterties, E. L. J. Am. Chem. Soc. 1977, 99, 8091−8093.(b) Brown, R. K.; Williams, J. M.; Sivak, A. J.; Muetterties, E. L. Inorg.Chem. 1980, 19, 370−374. (c) Ricci, J. S.; Koetzle, T. F.; Goodfellow,R. J.; Espinet, P.; Maitlis, P. M. Inorg. Chem. 1984, 23, 1828−1831.(9) The tetrarhodium complex [Rh(1,5-COD)(μ-H)]4 has also beentested as a precatalyst for toluene hydrogenation after adsorbing itonto silica or onto palladium supported on silica.9a The hydrogenationof carbon dioxide began with [Rh(1,5-COD)(μ-H)]4 and employedPh2P(CH2)4PPh2.

9b,c (a) Stanger, K. J.; Tang, Y.; Anderegg, J.;Angelici, R. J. J. Mol. Catal. A: Chem. 2003, 202, 147−161. (b) Gassner,F.; Dinjus, E.; Grls, H.; Leitner, W. Organometallics 1996, 15, 2078−2082. (c) Leitner, W.; Dinjus, E.; Gassner, F. J. Organomet. Chem.1994, 475, 257−266.(10) Alley, W. M.; Hamdemir, I. K.; Wang, Q.; Frenkel, A.; Li, L.;Yang, J. C.; Menard, L. D.; Nuzzo, R. G.; Ozkar, S.; Johnson, K. A.;Finke, R. G. Inorg. Chem. 2010, 49, 8131−8147.(11) Alley, W. M.; Girard, C. W.; Ozkar, S.; Finke, R. G. Inorg. Chem.2009, 48, 1114−1121.(12) By definition, Ziegler-type hydrogenation catalysts are formedfrom a non-zero-valent group 8−10 transition-metal precatalyst suchas [Ir(1,5-COD)(μ-O2C8H15)]2 plus a trialkylaluminum cocatalystsuch as AlEt3. For a review of the ∼50-year-old literature on Ziegler-type hydrogenation catalysts, see: Alley, W. M.; Hamdemir, I. K.;Johnson, K. A.; Finke, R. G. J. Mol. Catal. A: Chem. 2010, 315, 1−27.(13) Johnson, K. A. Polym. Prepr. 2000, 41, 1525−1526.(14) (a) Mondloch, J. E.; Finke, R. G. J. Am. Chem. Soc. 2011, 133,7744−7756. (b) Mondloch, J. E.; Wang, Q.; Frenkel, A. I.; Finke, R. G.J. Am. Chem. Soc. 2010, 132, 9701−9714. (c) Bayram, E.; Zahmakiran,M.; Ozkar, S.; Finke, R. G. Langmuir 2010, 26, 12455−12464.(d) Watzky, M. A.; Finney, E. E.; Finke, R. G. J. Am. Chem. Soc. 2008,130, 11959−11969. (e) Ott, L. S.; Finke, R. G. J. Nanosci. Nanotechnol.2008, 8, 1551−1556. (f) Ott, L. S.; Campbell, S.; Seddon, K. R.; Finke,R. G. Inorg. Chem. 2007, 46, 10335−10344. (g) Ozkar, S.; Finke, R. G.J. Organomet. Chem. 2004, 689, 493−501.(15) Alley, W. M.; Hamdemir, I. K.; Wang, Q.; Frenkel, A.; Li, L.;Yang, J. C.; Menard, L. D.; Nuzzo, R. G.; Ozkar, S.; Yih, K. H.;Johnson, K. A.; Finke, R. G. Langmuir 2011, 27, 6279−6294.(16) SADABS; Bruker Analytical X-Ray System, Inc.: Madison, WI,1999.(17) (a) Sheldrick, G. M., SHELXTL,, version 6.14; BrukerAnalytical X-Ray Systems, Inc.: Madison, WI, 1999. (b) A shorthistory of SHELX″: Sheldrick, G. M. Acta Crystallogr. 2008, A64,112−122.(18) Newville, M. J. Synchrotron Radiat. 2001, 8, 322−324.(19) Three main differences between Muetterties’ original procedureand Bonnemann’s synthesis are as follows: (i) Muetterties’ procedurestarts with 0.02 mmol of [Rh(1,5-COD)(μ-Cl)]2 versus 12.2 mmol inBonnemann’s synthesis; (ii) Muetterties and co-workers started withK[HB(O-i-C3H7)3] in THF or EtLi in toluene, whereas the hydridesource in Bonnemann’s synthesis is Na[BEt3H]; (iii) the initial[Rh(1,5-COD)(μ-Cl)]2 and hydride source are mixed immediately at−78 °C in Muetterties’ original synthesis versus over 9 h at roomtemperature in Bonnemann’s synthesis.3a,c

(20) Muetterties’ original [Rh(1,5-COD)(μ-H)]4 synthesis3a reports

50−60% yield when starting with 0.02 mmol of [Rh(1,5-COD)(μ-

Cl)]2. We have obtained a similar yield (∼50%) using the sameprocedure but when starting with 1.49 mmol of [Rh(1,5-COD)(μ-Cl)]2. However, using Bonnemann’s procedure,3c we obtain a loweryield (∼50%) than those reported by workers (74%), although ourscale of starting with 2 mmol of [Rh(1,5-COD)(μ-H]4 (vs 12.2 mmolin Bonnemann’s study) is one likely reason for our somewhat loweryield in that synthesis done purely as a control reaction.(21) Hampden-Smith and co-workers3b synthesized [Rh(1,5-COD)(μ-H)]4 by mixing 1 mmol of [Rh(1,5-COD)(μ-Cl)]2 and 2 mmol ofLiBEt3H at 0 °C and obtained a 52% yield. See this reference, as wellas ref 20, for a comparison to Muetterties’ and Bonnemann’ssyntheses.(22) The hydride sources K[HB(O-i-Pr)3] or EtLi used byMuetterties and co-workers3a were replaced by LiBEt3H in ourstudy because of the commercial unavailability of K[HB(O-i-Pr)3] andbecause Hampden-Smith and co-workers successfully used LiBEt3H intheir [Rh(1,5-COD)(μ-H)]4 synthesis.

3b An attempted synthesis usingEtLi was also performed as part of the present work but failed, asdetailed in the Supporting Information.(23) (a) The addition of free 1,5-COD is one key step that differsfrom Muetterties’ original synthesis and which results in a ∼78-foldhigher yield synthesis of [Ir(1,5-COD)(μ-H)]4. Another key differencefrom Muetterties’ original synthesis is that dark green [Ir(1,5-COD)(μ-H)]4 is kept in a THF solution until the filtration step (whereasMuetterties' synthesis of the Rh congener evaporates the volatilesunder vacuum and then extracts [Rh(1,5-COD)(μ-H)]4 usingpentane). (b) Additionally, an impurity detectable as a 12.82 ppmpeak in the 1H NMR is present in the black powder. That impurity canbe reduced by (i) washing the black powder with larger amounts ofdeoxygenated acetone in a drybox (a total of 250 mL vs 15 mL), (ii)increasing the stirring time from the initial 30 min to 24 h after theaddition of excess 1,5-COD, or (iii) passing a concentrated [(Ir(1,5-COD)(μ-H)]4 solution in THF through a glass filter (i.e., without theaddition of acetone). See the Experimental section of the SupportingInformation for further details and the 1H NMR spectra (Figures S1and S2 in the Supporting Information).(24) Similar to what has been found for [Ir(1,5-COD)(μ-H)]4, fourshort and two long M−M bonds are observed in the crystal structureof the Rh analogue. The four short bonds [2.802 (0.001) Å] wereassigned to bridging hydride Rh−H−Rh groups,3a consistent in ageneral way with the prior literature in which Rh−Rh bond distancesvarying between 2.610 (0.005) and 2.856 (0.008) Å have beenassigned to Rh−H−Rh groups.9a,b On the other hand, the long Rh−Rh distances in [Rh(1,5-COD)(μ-H)]4 of 2.971 (0.001) Å are longerthan the literature values of either singly bonded Rh−Rh (2.62−2.80Å9b) or Rh−H−Rh distances. (a) Brown, R. K.; Williams, J. M.; Sivak,A. J.; Pretzer, W. R.; Muetterties, E. L. Inorg. Chem. 1980, 19, 370−374. (b) Day, V. W.; Fredrich, M. F.; Reddy, G. S.; Sivak, A. J.; Pretzer,W. R.; Muetterties, E. L. J. Am. Chem. Soc. 1977, 99, 8091−8093.(25) The Ir−Ir bond distances for Ir−H−Ir linkages in Irn (n = 2−4)complexes of between 2.703 and 3.290 Å have been reported:(a) Heinekey, D. M.; Fine, D. A.; Barnhart, D. Organometallics 1997,16, 2530−2538. (b) Fujita, K.; Nakaguma, H.; Hanasaka, F.;Yamaguchi, R. Organometallics 2002, 21, 3479−3757. (c) Bau, R.;Chiang, M. Y.; Wel, C. Y.; Garlaschelli, L.; Martinengo, S.; Koetzle,T. F. Inorg. Chem. 1984, 23, 4758.(26) The Ir−C distances in [Ir(1,5-COD)(μ-H)]4 are similar to theIr−diene distances reported for26a [Ir4(CO)5(C8H12)(C8H10)] (2.10−2.24 Å) and to those previously observed in11 [(1,5-COD)Ir(μ-O2C8H15)]2. (a) Stuntz, G. F.; Shapley, J. R.; Pierpont, C. G. Inorg.Chem. 1978, 17, 2596−2603.(27) Similar M−M−M angles [54.93(11) and 62.53(5) Å] have beenseen previously for27a [(C5Me5)Rh(μ3-H)]4

2+. Table S2 in theSupporting Information contains a comparison of M−M−M bondangles of tetrahedral and butterfly-shaped M4 complexes.

8c

(28) Many Ir4 complexes such as [Ir4(CO)5(C8H12)2(C8H10)],[Ir4H10(PCy3)4(C9H11N)2](PF6)2, [(η

5-C5Me5)IrH]4(BF4)2, and tert-butylcalix[4]arene(OPr)3(OCH2PPh2)-Ir4(CO)11 have been success-fully characterized using MS. (a) Xu, Y.; Celik, M. A.; Thompson,

Inorganic Chemistry Article

dx.doi.org/10.1021/ic2026494 | Inorg. Chem. 2012, 51, 3186−31933192

A. L.; Cai, H.; Yurtsever, M.; Odell, B.; Green, J. C.; Mingos, D. M. P.;Brown, J. M. Angew. Chem., Int. Ed. 2009, 48, 582−585. (b) Cabeza,J. A.; Nutton, A.; Mann, B. E.; Brevard, C.; Maitlis, P. M. Inorg. Chim.Acta 1986, 115, L47−L48.(29) Tortorelli, L. J.; Flowers, P. A.; Harward, B. L.; Mamantov, G.;Klatt, L. N. J. Organomet. Chem. 1992, 429, 119−134.(30) Marshall, J. L.; Stobart, S. R.; Gray, H. B. J. Am. Chem. Soc. 1984,106, 3027−3029. (b) Rodman, G. S.; Mann, K. R. Inorg. Chem. 1988,27, 3347−3353.(31) For comparison, this −2.89 ppm chemical shift is considerablydownfield compared, for example, to the hydride peak for [Rh(1,5-COD)(μ-H)]4 at ca. −12 ppm. It is also downfield from the hydridesignals of previously synthesized tetrahydridotetrairidium complexessuch as [Ir(CO)(PPh3)(μ-H)H]4 or [Ir(η-C5Me5)(μ-H)]4(BF4)2appearing between −12.84 and −18.89 ppm.4,5 For a broadercomparison, the 1H NMR spectra of various complexes containingM−H−M (M: Ir−Rh or Re) face- or edge-bridging hydrides containhydride peaks between −4.30 and −24.00 ppm.31a−e,25a,b Hence,although one could speculate on the origins of this downfield, thehydride chemical shift (using, for example, the Ramsey shielding/deshielding equation31f or trying to take into account the increasedelectron density on Ir suggested by XANES), we choose not tospeculate in the absence of a good wave function and subsequentmolecular orbital calculation for [Ir(1,5-COD)(μ-H)]4. Note here thatthe needed computations would best come af ter definitive location ofthe hydrides by neutron diffraction, studies that are planned.(a) Vaartstra, B. A.; Cowie, M. Organometallics 1990, 9, 1594−1602.(b) Hattori, T.; Matsukawa, S.; Kuwata, S.; Ishii, Y.; Hidai, M. Chem.Commun. 2003, 510−511. (c) Mueting, A. M.; Boyle, P. D.; Wagner,R.; Pignolet, L. H. Inorg. Chem. 1988, 27, 271−279. (d) Fryzuk, M. D.Organometallics 1982, 1, 408−409. (e) Johnson, J. R.; Kaesz, H. D.Inorg. Synth. 1978, 18, 60−62. (f) Drago, R. S. Physical Methods forChemists; Surfside Scientific Publishers: Gainesville, FL, 1977.(32) (a) Adams, C. J.; Anderson, K. M.; Charmant, J. P. H.;Connelly, N. G.; Field, B. A.; Hallett, A. J.; Horne, M. Dalton Trans.2008, 2680−2692. (b) Brown, M. D.; Levason, W.; Reid, G.; Webster,M. Dalton Trans. 2006, 4039−4046. (c) Browning, J.; Bushnell, G. W.;Dixon, K. R.; Hilts, R. W. J. Organomet. Chem. 1993, 452, 205−218.(d) Zotto, A. D.; Costella, L.; Mezzetti, A.; Rigo, P. J. Organomet.Chem. 1991, 414, 109−118.

Inorganic Chemistry Article

dx.doi.org/10.1021/ic2026494 | Inorg. Chem. 2012, 51, 3186−31933193